Research On Control Methods Of Power Quality Synthetical Compensator

Unified Power Quality Conditioner (UPQC) is a new power electronic equipment with excellent performance, which can synthetically compensate familiar voltage and current quality problems of power system. However, there's no control method which is very suitable for UPQC.In the preceding work of our project group, two control methods of UPQC based on a unified math model that are synchronal deadbeat control and directly robust control are mentioned. This paper simplifies the math model first, and then researches on the improvement of these two methods based on the simplified model.Aiming at the flaw of deadbeat control that may be easily influenced by model parameters, system parameters identification online is used. Theoretical analysis and emulation result show that this method makes controller collaborate with model parameters well and keeps the rapid following characteristic of deadbeat control. Aiming at the flaw of directly robust control that can not dynamically compensate, an H∞control method based on mixed sensitivity design is mentioned. This method acquires good performance of disturbance restraining and robust stability by revising frequency characteristic of sensitivity function. It receives better compensating effect than directly robust control.In addition, neural network control technology is tried to apply in UPQC. A neural network control method which is in the form of uniform measure and independent control is mentioned, which can get better effect of load disturbance resistance than routine methods. And a neural network self-adaptively PID control is conceived. This method can not get good effect at present, but it can break the limitation of system model so that it has good foreground.This paper basically realizes routine control of UPQC through designing hardware circuit and software program for UPQC and analyzing the measured experiment curve.